BORN’S CONDITIONS ON THE WAVE FUNCTION

Link to: physicspages home page.

To leave a comment or report an error, please use the auxiliary blog.
Post date: 9 Jan 2021.
Max Born’s best known contribution to quantum mechanics was his pro-
posal that the wave function, or rather its square modulus, should be inter-
preted as the probability density for finding the system in a given state at a
given time. However, he also proposed four conditions on the wave func-
tion which are used in finding many solutions of the Schrödinger equation.
Remarkably few textbooks refer to these conditions, or at least, if they do,
they don’t credit Born as their originator.
As always, it’s useful to write down the Schrödinger equation (in one
dimension) so we can see how Born’s conditions fit in.

h̄2 ∂ 2 Ψ ∂Ψ
− 2
+ V (x, t)Ψ = ih̄ (1)
2m ∂x ∂t
Born’s conditions to be imposed on the wave function Ψ(x, t) are:
(1) The wave function must be single valued. This means that for any
given values of x and t, Ψ(x, t) must have a unique value. This
is a way of guaranteeing that there is only a single value for the
probability of the system being in a given state. Actually, if Ψ is a
proper mathematical function, it will satisfy this requirement auto-
matically, since one condition all functions must satisfy is that they
are single-valued. The most common example of a ’function’ (that
isn’t really a function) encountered by most undergraduate math
students is the inverse sine (or arcsin), which gives the angle cor-
responding to a particular value. Thus any multiple of π has a sine
of 0, so in principle, the inverse sine could give any multiple of π as
its value and thus it seems this is a multi-valued function. However,
in practice, only the range from 0 (inclusive) to π (exclusive) is used
in the ’proper’ arcsin function.
(2) The wave function must be square-integrable. In other words, the
integral of |Ψ|2 over all space must be finite. This is another way of
saying that it must be possible to use |Ψ|2 as a probability density,
since any probability density must integrate over all space to give
a value of 1, which is clearly not possible if the integral of |Ψ|2 is
1
 BORN’S CONDITIONS ON THE WAVE FUNCTION 2

infinite or zero. One consequence of this proposal is that Ψ must

tend to 0 for infinite distances.
(3) The wave function must be continuous everywhere. That is, there
are no sudden jumps in the probability density when moving through
space. If a function has a discontinuity such as a sharp step upwards
or downwards, this can be seen as a limiting case of a very rapid
change in the function. Such a rapid change would mean that the
derivative of the function was very large (either a very large posi-
tive or negative number). In the limit of a step function, this would
imply an infinite derivative. Since the momentum of the system is
found using the momentum operator, which is a first order deriva-
tive, this would imply an infinite momentum, which is not possible
in a physically realistic system. Such an infinite derivative would
also violate condition 4.
(4) All first-order derivatives of the wave function must be continuous.
Following the same reasoning as in condition 3, a discontinuous
first derivative would imply an infinite second derivative, and since
the energy of the system is found using the second derivative, a
discontinuous first derivative would imply an infinite energy, which
again is not physically realistic.

Having stated Born’s conditions, however, we need to note that several sys-
tems commonly studied in introductory quantum mechanics courses do vi-
olate one or more of them. For example, the ’particle in a box’ system
is composed of a particle moving in a box with infinitely high sides, rep-
resented by a potential function that is zero in a limited area, and infinite
outside this area. In such a system, the third condition (continuity of the
wave function) is imposed to find a solution, but the solution so found vio-
lates the fourth condition, in that the derivative of the wave function is not
continuous at the boundary of the box.
The particle in a box is clearly not a physically realistic system, how-
ever, since there is no known physical mechanism which can generate an
infinitely deep potential well. Despite that, the system is still useful as a
model of some real-life situations in which a particle is found in a very
deep well. And of course the particle in a box provides a relatively pain-
less way of introducing many of the concepts of quantum mechanics, so is
useful heuristically, even if it’s not entirely realistic.
It’s rather odd that most textbooks gloss over these conditions (well, they
usually mention the square integrable one since it’s essential for using the
wave function as a probability density) by simply stating that the wave func-
tion and its derivatives are required to be continuous, without explaining
why.
 BORN’S CONDITIONS ON THE WAVE FUNCTION 3

In the online MIT course, Barton Zwiebach discusses the requirements

that the wave function ψ (x) and its first derivative ψ 0 (x) are both contin-
uous, and gives examples of when these rules can be violated. We start by
looking at the time-independent Schrödinger equation in one dimension:

h̄2 d2
− ψ (x) = (E − V (x)) ψ (x) (2)
2m dx2
Basically, the restrictions on ψ arise from restrictions on the types of
potential function V (x) that we allow into the theory. We’ve seen examples
in the blog of an unbounded continuous potential (the harmonic oscillator),
a potential with finite discontinuities (the finite square well), a potential with
an infinite wall (sometimes called a hard wall: the infinite square well), and
one or more singularities, represented by delta functions. Generally, these
are the only types of potential that we will consider, so that, for example, we
don’t allow derivatives of delta functions or powers of delta functions. We
also don’t allow pathological potentials (such as the Weierstrass function,
which is continuous everywhere, but differentiable nowhere).
These conditions lead to the requirement that ψ (x) must be continuous
everywhere, although it may have a few isolated points where it’s not dif-
ferentiable. For example, the wave function for a delta function well has
a peak at the location of the delta function where it is not differentiable.
We require the first derivative ψ 0 (x) to be continuous everywhere except
at points where ψ (x) itself is not differentiable; at these points, ψ 0 (x) is
allowed to have a finite discontinuity (that is, a step up or down). This
happens with the delta function potential.
At points where ψ 0 (x) has a step discontinuity, its derivative ψ 00 (x) is a
delta function, since the derivative of a step function gives rise to a delta
function. This behaviour must be allowed in order for 2 to make sense.
d2
The second derivative dx 2 ψ (x) must contain the behaviour of the potential
due to the presence of V (x) on the RHS of this equation. If the potential
d2
contains a delta function then so must dx 2 ψ (x). However, since we’re not
allowing V (x) to contain derivatives of the delta function, then, working
backwards from ψ 00 (x), ψ 0 (x) cannot contain delta functions which means
that ψ (x) cannot contain any step functions (discontinuities). Thus ψ (x)
must be continuous.
There are basically four types of potential that we are allowed to consider:
(1) V (x) is continuous everywhere. That means that the RHS of 2 is
also continuous everywhere, so ψ 00 (x) must also be continuous ev-
erywhere. Again, working backwards, this means that ψ 0 and ψ
must also be continuous everywhere.
 BORN’S CONDITIONS ON THE WAVE FUNCTION 4

(2) V (x) contains finite discontinuities (jumps or steps). This means

that ψ 00 contains discontinuities, which in turn means that ψ 0 must
be continuous with a finite number of points where its derivative
(ψ 00 ) is discontinuous.
(3) V (x) contains delta functions. This is the case considered above,
leading to ψ 00 containing delta functions, ψ 0 containing discontinu-
ities and ψ being continuous.
(4) V (x) contains one or more hard walls, as in the infinite square well.
The wave function ψ is zero everywhere the potential is infinite, so
ψ 0 = 0 also in this region. However, as in the case of the infinite
square well, ψ 0 6= 0 just inside the wall, so ψ 0 is discontinuous at the
wall boundary. However, ψ itself is still continuous. Looking at 2,
it’s a bit difficult to interpret what happens to ψ 00 outside the wall
(where V = ∞), since the RHS is the product of ∞ with zero. To be
proper about it, we really shouldn’t consider what happens outside
the wall at all, since there is always zero probability that the particle
will ever be found there anyway.
R EFERENCES AND FURTHER READING
(1) Griffiths, David J. (2005), Introduction to Quantum Mechanics, 2nd
Edition; Pearson Education, Chapter 2. (Doesn’t mention Born by
name, but does use some of the conditions.)
(2) Zwiebach, Barton, Online course Mastering Quantum Mechanics
Part 1: Wave Mechanics. Archive available here.
P INGBACKS
Pingback: Infinite square well
Pingback: Delta-function well - bound state
Pingback: Finite square well - bound states, even wave functions
Pingback: Finite square well - bound states, odd wave functions
Pingback: Double delta function well
Pingback: WKB approximation